The investigators proposed to discover, through computation, lead compounds in the development of anti-AIDS drugs. Specific systems include HIV reverse transcriptase and HIV integrase. In addition, the investigators will attempt to refine their existing leads for HIV reverse transcriptase, HIV-integrase and HIV-RNA/TAR using DOCK and molecular dynamics-based methods coupled with structures if drug- receptor complexes. Aid will be provided to develop third generation versions of dominant negative strategy against HIV protease. Computational tools available for drug discovery and drug design will be developed and improved, with emphasis on speed enhancements, conformational searching, improved energy calculations combinatorial calculations, and anti-resistance strategies. Focus will be on an integrated approach to intensive optimization of the most important molecules using the following steps: (a) Aqvisits Linear Interaction Analysis (LIA) Method to suggest the most promising templates and the most favorable location of these templates in the macromolecule site; (b) Radmer's PROFEC methods coupled with the fragment and geometric search features of BUILDER, to suggest template modifications to improve binding; (c) Continued development of free energy calculations on a collection of the most promising substitutes of the most promising templates.